We hear a fair bit about the theory-ladenness of data, but little about the reverse.
Forgetfulness of the sometime theory-laden character of data is a complaint we make against the scientist who takes her data to offer more by way of confirmation of her theory than they possibly can. Her data can't play this role because their initial observation and gathering may itself have been guided by the theory in question.
We are also sometimes invited to acquiesce in the inevitability of this situation. Perhaps through some kind of conflation of theory and data with another dualism the terms of which, allegedly, are only ever intelligibly present when co-present - i.e. with thoughts or concepts and sensory experiences - the stronger idea seems to be that the concept of non-theory-laden data is necessarily but a fiction.
The philosophical claim sounds fishy to me and I'd rather stick with the more modest claim that sometimes a scientist's data is more theory-laden than she realises or would like to admit, and therefore that her offering it as support for her theory is more unconsciously narcissistic than she might care to acknowledge.
This post, however, is about the reverse predicament. My claim here is that theories are not always apt to offer explanations of the data they theorise not because the comprehension of that data is already theory-driven, but because the comprehension of the theory is already data-driven.
The claim is, in short, that we don't grasp what it means to say of someone that - to pick a random example - they have an unresolved oedipal conflict unless they behave like this --> or like that -->. The claim is that whilst our grasp of these theoretical terms ('oedipal conflict') might look like the kind of thing that could be so much as understood in abstraction from our experience of the relevant 'data' (i.e. in abstraction from situated observations of human behaviour), this in fact is an illusion. We - especially we psychologists - pick up and flesh out and create the meaning of our terms in and through an ongoing immersion in the clinical situations we inhabit. We are, one way or another, given various paradigm cases of what we're talking about. We get a feel for what is going on, a gestalt self-organises. What the theoretical terms do, I'm claiming, is rather often to 'honour' (offer recognition to) the gestalts which spontaneously organise in experience. Where by 'spontaneously organise' I do not mean: organise independently of any learning or any ongoing sensitive immersion in some or other situation. I mean, instead: organise independently of the constraining and inspiring prompts of the theory in question.
We tend to think that theoretical terms are intelligible in pure cogitation - i.e. in abstraction from the data. We then imagine, in a Kantian spirit, that if our data seem to fit the theory rather nicely, this is because we have tacitly gathered them through the matrix of our theoretical apprehension. I hope it's clear how big a 'narcissist' sign this also has written in marker pen across its own forehead. For notice how extraordinarily confident it is in supposing that the disengaged mind can know what on earth it's going on about without its meanings yet being in-formed by a living participation in relationships with others and with the world about us. As if the relation of meaning to use was simply from the former to the latter. As if ostension and conditioning and the whole caboodle of praxis was some rather secondary matter. As if we could just arrogate to our own comprehensions a grasp of the meaning of our terms without that hanging on any kind of praxical demonstration.
Sometimes of course we can use non-ostensive definition to specify the meaning of our terms. And in such cases the new terms are intelligible without our already having participated in new sensorimotor, perceptual-gestalt-inviting, world-disclosing, habits which the terms honour. Of course this will usually just push the dependency on praxis back a stage - to the grasp of the meaning of the other terms in which the definition is given. My claim, however, is that such situations are rather atypical. Atypical, at least, when we're thinking of fundamental theoretical notions in, say, psychodynamic psychology - such as repression, projection, sublimation, etc. A breezy talk of 'defence mechanisms', a too-ready availability of mentalistic metaphors to frame an over-confident elaboration of such mechanisms (of, you know, the 'he transfers his own inner representations to the interior of the person's mind he is himself representing' sort - of a sort which makes the discourse look far more conceptually independent of clinical experience than it really is), disguises from us the overwhelming significance of intuition and immersed learning in grounding our grasp of the relevant concepts. Or, disguises from us the internal relation of theory to data - the data-ladenness of our theory - or encourages the theorisation of that internal relation simply in terms of the theory-ladenness of our data. Sure, sometimes a theory may help us get our eye in. What I'm making some noise about here, however is the significance, for our grasp of the meaning of our theory, of having our eye already in.